Biotechnology Letters 26: 6770, 2004.

2004 Kluwer Academic Publishers. Printed in the Netherlands.

67

Spectrophotometric method for determining gibberellic acid in

fermentation broths
Julio Berros , Andres Illanes & German Aroca
Escuela de Ingeniera Bioqumica, Pontificia Universidad Catolica de Valparaso, General Cruz 34, Valparaso,
Chile
Present address: Escuela de Ingeniera en Biotecnologa, Universidad San Sebasti
an, Cruz 1577, Concepcion,
Chile
Author for correspondence (Fax: +56 (32) 273803; E-mail: [email protected])
Received 6 August 2003; Revisions requested 3 September 2003; Revisions received 3 November 2003; Accepted 4 November 2003

Key words: Gibberella fujikuroi, gibberellenic acid, gibberellic acid

Abstract
A novel method for the quantitative determination of gibberellic acid in fermentation broths has been developed.
It is based on the kinetic of the reaction of conversion of gibberellic acid to gibberellenic acid. The method
is simple, reliable, faster than most of methods known, and free of the interferences which commonly affect
spectrophotometric methods currently in use. Its threshold sensitivity is 0.1 g and its accuracy is greater than 97%
for concentrations of gibberellic acid ranging from 0.1 to 1 g l1 .

Introduction
Gibberellic acid (GA3 ) is a naturally occurring plant
growth regulator which may cause a variety of effects
including stem elongation. Methods for gibberellic
acid (GA3 ) quantification were initially based on the
measurement of growth changes of different plant
organs and parts promoted by GA3 . Bioassays are
currently used when high levels of specificity and
sensitivity are required, but they are not suitable for
monitoring GA3 during fermentation, where results
are promptly required. For this purpose, simpler and
faster methods, such as colourimetric, spectrophotometric, and fluorimetric are preferred. Nevertheless,
these methods have the disadvantage that the sample
requires pre-treatment for removing interfering substances present in the fermentation broth.
The spectrophotometric method proposed by Holbrook et al. (1961) is one of the simplest and most
widely used methods for GA3 determination in fermentation studies (Kahlon & Malhotra 1986, Hollmann et al. 1995, Tomasini et al. 1997). Briefly,
this method consists in the addition of HCl to the
sample to then measure the absorbance at 254 nm after

75 min. During this time, the absorbance rises up to a

maximum value and then it falls slowly. This is explained by Holbrook et al. (1961) as the combined
effect of two chemical reactions: (i) GA3 conversion
to gibberellenic acid (GE) and (ii) GE decomposition
to gibberic and allogibberic acids. These ones do not
absorb in the UV range. The latter reaction is not pHdependent and occurs spontaneously. Broth samples
require pre-treatment and solvent extraction of GA3 .
The production of GA3 by fermentation using
Gibberella fujikuroi have been extensively studied; for
comprehensive reviews see Jefferys (1970) and Kumar
& Lonsane (1989). A high improvement in the yield
of GA3 is obtained by adding corn steep liquor to
the culture medium, however, this media formulation
contains several compounds that absorb strongly in the
UV range, changing the absorbance continuously during the fermentation. For these reason, the use of nondefined media makes the application of the method
proposed by Holbrook et al. almost impractical to
follow a fermentation.
In this work we propose a method for the quantification of GA3 in fermentation broths that avoids
the sample pre-treatment procedure proposed by Hol-

68
brook et al. (1961) and takes advantage of the fact that
the initial rate of conversion of GA3 to GE is lineal
at least up to 2 min of reaction when HCl is added to
the sample, and the slope of the conversion line can
be related to the initial concentration of GA3 in the
sample.

Materials and methods

GA3 production by fermentation
GA3 was produced by fed-batch fermentation using
the fungus Gibberella fujikuroi ATCC 12616. The fermentation was carried out in a 3 l bioreactor (New
Brunswick, NJ) at 30 C, with 2 l culture medium
in the batch stage of the following composition: glucose 20 g l1 , corn steep liquor 25 g l1 , (NH4 )2 SO4
3.3 g l1 , KH2 PO4 0.5 g l1 , K2 SO4 0.2 g l1 . After
inoculation the fungus was allowed to growth for 24 h
then the culture was fed at constant rate of 20 mg
per h during the next 7 d where the concentration
of GA3 reached its maximum (0.66 g l1 ) under the
given culture conditions. Aeration rate was 1 vvm and
the agitator speed 400 rpm. Biomass of broth samples
was measured by dry weight as described in Kahlon &
Malhotra (1986).
Pre-treatment of samples
Broth samples were centrifuged to remove biomass.
An aliquot expected to contain between 2 and 6 mg
of GA3 was transfer to a 100 ml separating funnel.
Water was added to make 10 ml. The pH of the
solution was adjusted between 1 and 2 using 0.1 M
HCl. Twenty ml of ethyl acetate was added and vigorously shaken for 60 s. The aqueous phase was then
transferred to a second separating funnel and the extraction procedure repeated by adding another 20 ml
ethyl acetate. The aqueous phase was then discharged
and the organic phase transfered to the first separating
funnel. The GA3 was re-extracted from ethyl acetate
with successive portions of 20, 15 and 10 ml phosphate buffer (pH 7.4), shaking each time for 60 s, and
combining each extraction in a 50 ml volumetric flask.
The volume of this flask was made up with phosphate
buffer.
Determination of GA3
A sample of 1 ml and 1 ml absolute ethanol were
placed in a 10 ml volumetric flask. HCl, 3.75 M,

Fig. 1. Gibberellenic acid formation from GA3 by adding 3.75 M

HCl. Absorbance in arbitrary units (AU). GA3 initial concentration
(g l1 ): (
) 0.1, (
) 0.2; () 0.4; () 0.6; () 0.8; (+) 1.

was added to the flask up to make 10 ml and then

vigorously mixed for 10 s. The absorbancy of the resulting solution was measured at 254 nm and recorded
at 20 s intervals for 2 min. Temperature was kept at
20 0.5 C during the process. A calibration graph
was then obtained by using standard GA3 solutions
prepared by dissolving 0.04 g pure GA3 in absolute
alcohol and diluted to 100 ml in a volumetric flask
with absolute alcohol.
Calculations
Each series of data obtained from spectrophotometric
measurement were fitted by linear regression analysis
using a computer software. The slopes obtained from
each series of determinations were analysed by linear regression, generating a calibration graph for the
determination of the concentration of GA3 .

Results and discussion

Figure 1 shows the linear response obtained when absorbance was plotted against time within 2 min after
adding HCl to the standard solutions of GA3 . The
linear regression coefficient (R) obtained was always
higher than 0.998. The absorbance reached a maximum value at 60 min and then it falls slowly. The
value of the slope of each line correlates with the initial
concentration of GA3 in the sample. Figure 2 shows
the relation between those slopes and the initial concentration of GA3 . The values of ri were obtained
by linear regression of each series of measurement

69

Fig. 2. Effect of GA3 initial concentration on the initial rate of

conversion of gibberellic acid into gibberellenic acid (calibration
curve).
Table 1. Concentrations of GA3 in prepared
solutions added to and measured by the proposed method in the culture media.
GA3 added
(g l1 )

GA3 measured
(g l1 )

Error (%)

0.2
0.4
0.6

0.2
0.39
0.59

0
2.5
1.6

at different GA3 concentration (Figure 1). This relation allows to determine the concentration of GA3
by determining the initial rate of GA3 decomposition
into GE. The dotted line shows the linear regression
(R = 0.9993).
Different buffer concentrations in the sample solution were tested in order to determinate the effect of
the buffer strength on the lineal relation above mentioned. These experiments were carried out to test the
potential effect of changes in buffer strength that occurs during the production of GA3 by fermentation.
No effect was found when samples of GA3 at 0.4 g l1
were prepared in 0.01, 0.05, 0.1 and 0.5 M phosphate
buffer at pH 5. This value was chosen because the pH
in the fermentation broth usually varies around this
value during the fermentation. Analogous results were
obtained using 0.2 and 0.6 g GA3 l1 .

of GA3 was measured using solution of GA3 prepared

with the culture medium. The results obtained are
shown in Table 1. The error obtained is slightly higher
than the one obtained with aqueous solution of GA3
and it is similar to the error reported by Holbrook et al.
(1961).
In order to validate the assay, GA3 produced by
Gibberella fujikuroi was measured by both the method
described here and Holbrook method. The final concentration obtained was 0.66 g l1 by the former
method and 0.65 g l1 measured by the latter. Similar results were obtained using samples obtained from
three different cultures. Pretreatment of the broths
samples was required only when Holbrooks method
was used.
Holbrook et al. (1961) studied the effect of a
number of possible interferences such as the presence of GE, isogibberellic, allogibberic, gibberic
acids, and Sumikis acid (5-hydroxymethylfuran2-carboxylic acid) in the samples, produced as
byproducts in the fermentation. None of these compounds exhibits changes in their absorption under the
experimental conditions. Nevertheless, Sumikis acid
absorbs strongly in the UV, having a maximum at
260 nm, and a long experimental procedure to quantify
the effect of this interference is proposed.
In the method proposed here, the effect of any absorbing compound present in the sample has not effect
on the initial rate of decomposition of GA3 to GE,
since its initial rate depends only on the initial concentration of GA3 . Thus, the method described here
does not require corrections to these interferences.
The range of GA3 concentrations studied was 0.1 to
1 g l1 . The sensitivity limit of the method is 0.1 g l1
of GA3 . The standard error obtained in three measurements was < 3% for concentration sample between
0.1 to 1 g l1 .

Acknowledgement
This work was funded by Pontificia Universidad
Catlica de Valparaso, Valparaso, Chile, Project
203.704/98

References
GA3 determination in culture media
In order to measure the effect of the composition of
the culture medium over the assay, the concentration

Holbrook A, Edge W, Bailey F (1961) Spectrophotometric method

for determination of gibberellic acid. Adv. Chem. Ser. 28: 159
167.

70
Hollmann D, Switalski J, Giepel S, Onken U (1995) Extractive fermentation of gibberellic acid by Gibberella fujikuroi. J. Ferment.
Bioeng. 79: 594600.
Jefferys E (1970) The gibberellin fermentation. Adv. Appl. Microbiol. 13: 283316.
Kahlon S, Malhotra (1986) Production of gibberellic acid by fungal
mycelium immobilised in sodium alginate. Enzyme Microb.
Technol. 8: 613616.

Kumar P, Lonsane B (1989) Microbial production of gibberellins:

state of the art. Adv. Appl. Microbiol. 34: 29139.
Tomasini A, Fajardo C, Barrios-Gonzlez J (1997) Gibberellic
acid production using different solid- state fermentation systems.
World J. Microbiol. Biotechnol. 13: 203206